Re: SQL Agent Questions

Chris,  Thanks for the response...this agent stuff is new to me so if you
could clear up a couple of more points, I would appreciate it.

1.   The size originally created for the transaction log piece of the
database is 1.5GB.  To get this to go directly to disk, we created a
storage pool size of 1.5GB; however, when we backed up the log, it only
saved 3.3MB of data.  Our original disk pool size was 500MB which should
have been more than sufficient.  Does the SQL agent use the total size of
the transaction log as its indicator of where to send the data (i.e. if the
disk pool isn't as large as the transaction log, will it try to go directly
to the next pool in the hierarchy?)

2.   When we do a scheduled backup of the transaction log, (I think the
command is sqldsmc /backincr .....), does this automatically truncate the
transaction log?  And just so I understand, the truncation only gets rid of
what's considered the 'inactive' part of the log, is that correct?

3.   Is the SQL not designed to restore individual tables, etc?  or do you
have to completely restore an entire database or individual transaction
logs?  I'm not sure what the best way is to implement the agent.
---------------------- Forwarded by Sharon Combs/Tulsa/IBM on 02/26/99
10:27 AM ---------------------------
10:27 AM ---------------------------


Chris Zaremba <zaremba AT US.IBM DOT COM> on 02/24/99 04:22:49 PM

Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>

To:   ADSM-L AT VM.MARIST DOT EDU
cc:    (bcc: Sharon Combs/Tulsa/IBM)
Subject:  Re: SQL Agent Questions





Sharon,

>Is there
>something different in the way the SQL Agent runs as compared to the
>regular ADSM Backup-Archive client???

Not in terms of how it binds objects to a management class


>Does the SQL Agent require that if
>the first entire server backup was done to tape that the remaining
>transaction logs are also required to go to tape?

No

>We want to be able to
>take the first full backup of the month offsite and keep the changes
>(i.e. the transaction logs) on disk.  What am I missing????

I would suggest you check the following:

1. Is the SQL agent using a management class that sends the backups
directly to tape?
    (Check the default management class for policy domain in which the node
is registered?)

2. Do you have a maxsize threshhold on the disk pool that is being
exceeded?  That is, are
     the incrementals large enough to bypass the disk pool?

3. Is there ample free space in the diskpool ?

Having said all that,  have you considered binding all log files to a
special management
class that sent them to a diskpool that didn't migrate?  thast would
require setting up a separate
pool for this purpose but could eliminate flipping migration threshholds...

Chris Zaremba
ADSM Agent Development
internet  zaremba AT us.ibm DOT com